Journal of Space Technology, Volume V, No.1, July 2015
Design and Development of Accelerated Life
Tester for Qualification of Batteries
Syed Muhammad Arsalan Bashira, Syed Kazim Hasan Zaidib, and Muhammad Zaheerc
Abstract—Accelerated life testing (ALT) of batteries is
performed to estimate the number of life cycles a battery can
withstand in any mission fulfilling the mission power
requirements. This test is destructive; therefore the battery cycles
are exhausted completely. For ALT, selective samples are used to
qualify the whole lot. Life of any space mission is dictated by
batteries used in the mission, therefore the qualification process
of batteries is quite critical.
The present paper elaborates the design and development of
Accelerated Life Tester for batteries (all types) according to the
requirements of the Society of Automotive Engineer (SAE)
standards. The setup is able to perform both charge and
discharge cycles simultaneously with programmable cutoff
voltage and current limits. Dedicated data acquisition system
(DAQ) is developed in LabView that is capable of receiving data
at high rates. Different sensors are used to monitor and control
the test. The recorded data include body temperature of
batteries, terminal voltages and current drawn. The real time
data is stored in the LabView data file for post processing and at
the same time it is shown as real time graphs for monitoring the
test through a graphical user interface (GUI). Qualification of
batteries is performed on selected samples in extreme test
conditions, including high temperature, a high discharge rate and
a high depth of discharge (DOD). The present setup provides
these test conditions for performing ALT of batteries. The GUI
controls every function of the test centrally, from test initiation to
test termination that makes the test setup fully automated.
Keywords—accelerated life test; cycle life tester; accelerated life
test for batteries; cycle life tester.
I.
INTRODUCTION
Accelerated life testing is of immense importance in
predicting the life of an electrical component in a system. For
power sources like batteries, accelerated life testing is vital for
predicting the useful functional life of batteries. In this paper,
we are presenting the design and development of an
accelerated life tester for batteries, which will analyze the
charge-discharge cycle of batteries. Major parameters
(including charge/discharge rate, cutoff voltages, temperature
thresholds, etc.) are soft programmed, and can be set
according to the specification sheet of the battery under test.
This test determines the number of life cycles a battery will
survive. This test is destructive as the battery cycles are
exhausted completely and hence the batteries cannot be used
after ALT completely so selective samples may be used to
qualify the whole lot.
In [1], operating temperature, charge rate and discharge rate
are the main factors that define the degradation in the
efficiency of high capacity batteries. In our design criteria, the
critical test parameters are same as discussed in [1]. In [2], the
authors developed a life model for lithium-ion battery
generally used in electric vehicles where they studied the
internal capacity fading mechanism of the battery under test.
The test can be of two types, selective life cycle test and a full
life cycle test. As their name suggests, in the selective life
cycle testing the battery under the test is analyzed for a few
cycles and the life cycle is extrapolated using various models,
while full life cycle testing estimates actual cycles a battery
can withstand till it reaches failure limits.
Various prediction models can be used to extrapolate the
results for life estimation, the basic parameters that dictate the
life of a battery are, operational temperature, the
charge/discharge rate and depth of discharge (DOD). The
exact cycle life estimation of batteries may take years in the
case of valve regulated lead acid (VRLA) batteries. As in [3]
authors analyzed the cycle life of lithium secondary batteries
along with FMEA (failure mode and effect analysis) and the
safety/abuse tests to access reliability of secondary batteries.
Also in [4] the authors discussed the life cycle failure rates of
VRLA batteries against the expected life of the batteries and
concluded that the average lifetime of VRLA batteries is in
between 13 to 15 years. In [5], the authors discussed the
relation of the cycle and calendar life. They used three diverse
temperature ranges to achieve the end of life (EOL) for a
certain battery sample. In our approach, there are four major
factors that define an accelerated life cycle test of a battery:
1-Battery type, 2-Charging cycle, 3-Discharging cycle, and 4Test conditions
A. Battery Type
Manuscript received March 20, 2015.
a
Assistant Manager, QA Electronics Dte, QA Electronics Division, Quality
Management Directorate General. Pakistan Space and Upper Atmosphere
Research Commission (SUPARCO), 75270 Karachi, Pakistan. Email:
smab1176@yahoo.com. Tel.: +92-345-2252898; fax: +92-21-32593223.
b
Manager, QA Electronics Dte, QA Electronics Division, Quality
Management Directorate General. Pakistan Space and Upper Atmosphere
Research Commission (SUPARCO), 75270 Karachi, Pakistan. Email:
zkashmiri@hotmail.com. Tel.: +92-321-3626268; fax: +92-21-32593223.
c
Manager, QA Electronics Dte, QA Electronics Division, Quality
Management Directorate General. Pakistan Space and Upper Atmosphere
Research Commission (SUPARCO), 75270 Karachi, Pakistan.
Email: zkashmiri@hotmail.com. Tel.: +92-321-3626268; fax: +92-2132593223.
Battery type is very vital, for heavy duty batteries the depth of
discharge (DOD) may be selected up to 100% but for standby
batteries like those used in vehicles; a low DOD must be
selected. Also in batteries with memory effect like, nickelcadmium (NiCd) batteries the DOD must be selected closer to
100% so to ensure that battery capacity is not compromised
because of the memory effect.
33
Journal of Space Technology, Volume V, No.1, July 2015
C. Discharging Cycle
Discharge rate is very critical parameter for power sources and
their accelerated life testing. This parameter dictates the
capacity of any power source. The quoted capacity of any
battery normally requires that its discharge current should be
0.1C. As the discharge rate increases the battery capacity
decreases therefore evaluation of any battery must be done
considering this fact. For a standard medium size lead-acid
battery of 150Ahr capacity, the discharge rate of 25.5A leads
to a rated capacity of 127.5Ahr. Thus, a discharge rate of
25.5A will guarantee 5 hours of backup time at ambient
conditions. The accelerated life test is done near to the
maximum discharge rate to simulate the adverse conditions.
As in [6], the authors devised a method for estimating the
lifetime of valve-regulated lead-acid (VRLA) batteries used in
float charge service, under a variable-temperature
environment. They concluded that by increasing discharge
current used in a cycle lifetime test, it is possible to shorten the
period required for the cycle lifetime test.
D. Test Conditions
Fig. 1.Functional block diagram of the life cycle tester for batteries.
1.
For complete ALT, batteries should be new and
unused.
2.
During the entire charging cycle, electrolyte
temperature must be maintained between 16°C and
43°C.
3.
Electrolyte strength: batteries shall be tested with the
electrolyte as supplied by the manufacturer.
4.
Charging and discharging current: the charging and
discharging current must be selected close to the
maximum rated current. Capacity measured must be
analyzed in accordance to the discharge current as
mentioned in the data sheet. The high discharge
current (higher than the maximum rated current) leads
to a lower battery capacity. The battery discharge time
can be calculated using Peukert's law as in (1)
B. Charging cycle
The charging cycle starts where the battery is provided with
external power to recharge the energy dissipated by the load.
The cycle is initiated as soon as the inward current starts
flowing. The charging current is selected as per specification
sheet of the battery; the charge rate in ALT is kept closer to
the max charge rate to simulate adverse conditions. The
charging cycle completes when the charging current reaches
0.02 x current of its rated value. The test conditions are
explained in later sections. Fig. 1 shows the functional block
diagram of the system with major components.
Fig. 2. Block diagram of the data acquisition system for ALT of batteries.
34
Journal of Space Technology, Volume V, No.1, July 2015
 C 
t = H

 IH 
A. Data Acquisition System Using the NI LabView
K
Data acquisition system (DAQ) is built on LabView interface
using serial port for sensor data logging and parallel port for
command and control. The functional block diagram of DAQ is
presented in Fig. 2. It functions in following steps:
Where,
t is actual discharge time in hours
H is rated discharge time in hours.
1.
Serial data are received from the com port at the rate
of 19200 bps (although it can accept any data rate in
accordance with the RS232 protocol).
2.
The next step is the frame identification and byte
separation. This is done by adding a frame identifier at
the beginning of every sequence. Acquired data is not
validated until a frame identifier is detected.
3.
After segregating the data, the next step is to perform
signal-conditioning. Scale factors for voltage, current
and temperature sensor readings are incorporated
along with required off-sets in the LabView interface.
4.
Real-time graphs of all the sensor reading are
displayed and the data is logged into LabView
database file for post processing.
5.
For data recording, update rate can be set as low as 50
ms.
6.
The decision making is based on the sensor data for
temperature ranges, cutoff voltages and charging
current. As any of the limits are reached the test is
stopped.
7.
A real-time clock is set as the test starts and the total
test time is shown and logged for post processing.
I is actual discharge current in ampere.
C is the rated capacity in ampere hours.
K is the dimension-less Peukert constant, usually its value
is from 1.1 to 1.3.
II.
ACCELERATED LIFE TESTER
The basic functional block diagram of an accelerated life tester
for batteries is illustrated in Fig. 1. Major components of the
tester are further explained in later subsections. A high
wattage power supply provides the charging current. Overload protection is ensured by including the electronic relay for
central control. Sensors for monitoring and controlling the test
are mounted on the battery under test and are connected via
the main board to a PC for data acquisition. The system is
designed to operate two samples simultaneously. Discharging
part of the tester consists of a programmable DC load, overload protection circuit, electronic relay control and battery
under test. The DC load is programmed as per required
discharge rate of the battery under the test. The test is initiated
and terminated via front panel software control.
The subsystem level description is as follows:
Fig. 3. Front panel of data acquisition system.
35
Journal of Space Technology, Volume V, No.1, July 2015
Filtering is done to remove any ripples that are generated due
to loading. The power lines are regulated and monitored for
any sudden distortion.
C. Sensor and Signal Conditioning
Fig. 4. Block diagram of the power distribution unit
Fig. 7. Block diagram of the power board.
The sensors are directly mounted on the battery under test. PT100 (Platinum Resistance Thermometer) is used as a
temperature sensor and is mounted directly to the battery just
below 25% of its height from the top. For terminal voltage
monitoring, the voltage lines are directly connected to the
battery. For current measurement hall sensors are used with an
accuracy of 0.1A. Fig. 5 shows the sensor and signal
conditioning card. All the sensors are fed directly to the main
board for further processing. For temperature sensors
compensated wires are used for exact measurement. Signal
conditioning is done to convert signals to desired signals so
that it may be fed directly to the controller. The voltage levels
are controlled in between 0V-5V.
D. Main System Board
Fig. 8.System under operation (charge and discharge cycle, 2
simultaneous samples).
8.
The test is controlled using the parallel port via
electronic relays.
Fig. 3 shows the GUI of DAQ while undergoing charge and
discharge cycle. The left column shows the visa settings
including the baud rate, data bits, parity control, stop bit (if
any) and flow control (if any). The central column shows
sensor readings for the charge and discharge cycle. The central
column top buttons are test initiation buttons. Real-time
graphs are shown in extra tabs named as charging and
discharging.
B. Power Distribution Unit
Fig. 4 shows the power distribution card where a single DC
supply voltage is converted into various power busses for
different cards. These conversions are DC-DC convertor based.
Fig. 5. Block diagram of sensors and signal conditioning unit
The main board consists of the functional blocks as illustrated
in the Fig. 6. Multiplexer (MUX) is used to multiplex the
sensor data. Controller has 8 analog ports and it can read 6
analog channels, but this is implemented to increase the
capability of the system. MUX16ET is basically a 16 to 1 wire
converter. This is then read by a microcontroller with its analog
port and after processing the MUX data; the controller
transmits the frame to serial port via level converter IC. The
channel addresses are also generated by the controller and
frame structure is also developed by the controller. The serial
communication baud rate is set by the controller and DAQ
software uses the same baud rate to receive the data.
E. Power Board
Power board is responsible for the actual charge and discharge
cycle initiation and termination. The block diagram is shown as
Fig. 7. The signal control is generated by the DAQ software
which acts as the test initiation or termination signal. This
signal is directly fed to an electronic relay which is responsible
to connect the power lines for charging and discharging cycle.
Circuit breaker is connected in a series for protection against
any fatal accident.
F. Water Bath and Test Setup
The test is performed in a temperature controlled environment
when the temperature is maintained within 20°C to 30°C. For
performing the test under extreme conditions, a water bath
along with heating control is required. This will decrease the
battery cycles and consequently it will result in early test
completion time. The battery is dipped to 75% of its volume
Fig. 6. Block diagram of the main system board
36
Journal of Space Technology, Volume V, No.1, July 2015
keeping the top layer and terminals out of the bath and the
temperature sensor is mounted just below the water level.
III.
EXPERIMENTAL RESULTS
The test is performed on three different samples with
specifications and brand names shown in the Table 1.
TABLE I.
SAMPLES TESTED FOR ALT OF BATTERIES
Sample 1
Sample 2
Sample 3
Sacred Sun
Sacred Sun
Leoch
Medium capacity VRLA
VRLA Gel
Maintenance free
SP12-150
6GFMJ-150
LP12-150
www.sacredsun.com/Upload/pr
oducts/pdf/SP12-150.pdf
www.sacredsun.com/Upload/pr
oducts/pdf/6GFMJ-150H.pdf
http://www.leoch.com/pdf/r
eserve-power/agm-vrla/lpgeneral/LP12-150.pdf
12V – 150Ahr
12V – 150Ahr
150Ahr @ 0.05C
150Ahr @ 0.1C
Max charge rate 45A
Max charge rate30A
Max discharge
rate1500A (5 sec)
Max discharge rate
1472A (3 sec)
Cutoff voltage 10.5V
Cutoff voltage 10.8V
Operational
temperature: -10°C-
Operational
temperature: -20°C-
45°C
45°C
200 rated life cycles for
100% DOD
700 rated life cycles for
100% DOD
Fig. 10. Charging voltage profile for cycle 6 of 3 samples under test
12V – 150Ahr
150Ahr @ 0.1C
Max charge rate 45A
Max discharge rate
1500A (5 sec)
Cutoff voltage 10.5V
Operational
temperature: 0°C40°C
200 rated life cycles
for 100% DOD
Fig. 12.Discharging voltage profile for cycle 6 of 3 samples under test
Fig. 9. Charging current profile for cycle 6 of 3 samples under test
Fig. 13.Capacity curve of the samples after 16 cycles
37
Journal of Space Technology, Volume V, No.1, July 2015
After 16 cycles, sample 2 has prominent result. The test is
ongoing till the capacity degradation of 20% is achieved.
IV.
Fig. 11. Discharging current profile for cycle 6 of 3 samples under test
Full system under operation is shown in Fig. 8. In Fig. 8,
the DAQ software shown on the monitor screen controls the
parameters and the electronics involved are placed on the right
side. As per manufacturer specifications, the discharge rate of
25.5A was selected for sample 1 and 2 and 25.8A for the
sample 3. These rates lead to a rated discharge time of 5 hours;
the batteries are tested until the battery capacity falls below
80% of its initial rated capacity. The charge rate is selected
same as discharge. The charging voltage is selected as per
battery specifications and for 12V batteries it is 14.8V±0.05V.
The charging current and voltage profiles of the three samples
(after testing of 16 cycles) are shown in Fig.9 and Fig. 10. For
sample 1, the constant current charging lasts less than that of
sample 2 and 3. For the same reason, sample 1 reaches the
constant state voltage charging earlier than sample 2 and 3.
The discharge cycle profiles (for both current and voltage)
are shown in Fig. 11 and Fig. 12. In Fig. 11 the current profile
is shown. Constant current discharging is shown where sample
3 has a rate of 25.8A and sample 1 and 2 are discharged at a
rate of 25.5A. Sample 2 has the highest discharge time
illustrating it has the highest capacity.
Fig. 12 shows the terminal voltage curve while the
discharging cycle is in progress. It is evident that the major
portion of the battery capacity is stored in the span of 13V11.5V. Fig. 13 shows the actual result of the capacity
degradation of the three samples it is evident that the capacity
of sample 1 is decreased more than the other two samples. The
capacity status of the samples after 16 complete cycles is
shown in Table II.
TABLE II.
CAPACITY STATUS AFTER 16 COMPLETE CYCLES
Sample 1
Sample 2
Sample 3
Initial Capacity:
Initial Capacity:
Initial Capacity:
107.3 Ahr
136.8 Ahr
130.1 Ahr
Capacity After 16
Cycles:
Capacity After 16
Cycles:
Capacity After 16
Cycles:
91.1 Ahr
136.1 Ahr
121.4 Ahr
Capacity degradation:
Capacity degradation:
Capacity degradation:
15.1%
0.5%
6.7%
APPLICATIONS
Applications are huge from power management of ground
vehicles to commercial application in battery analysis of
power sources to be incorporated in space missions. As the
battery life is a critical parameter of any product so this test
can be used for battery selection for any particular mission.
Furthermore, applications where the dynamic use of the
battery is required, where there is less time between the charge
and discharge cycle; accelerated life tester may be used for
selecting a suitable type of battery. In consumer applications
for selecting better power sources as in [7] the authors carried
out research to make the results of an accelerated life test of
the cycle-life products more accurate. A more suitable model
is required for predicting the cycle-life in [8] the authors
developed a cycle-life estimation model for lithium-ion
batteries based on limited cycle data. This model was a
straight line estimation model and it had an error of 40%.
Hence exponential modeling must be done in the future.
V.
CONCLUSION
The paper discussed the design and development of an
accelerated life tester. Partial test results were discussed and
the applications are in various fields. The tester has a fast
response of 50 ms per frame acquisition. The data are logged
and are post processed. The setup is fully automated and
requires initial test settings. The system is robust as it is tested
for a continuous operation of 8 hour charging cycle and 5 hour
discharging cycle for a continuous 3-month period. Data
accuracy of 10mV is also achieved.
The future work includes the analysis and modeling of the
life cycle using accelerated life testing results. This will
include the formulation of a mathematical model that may
predict the results with a certain level of confidence. As in [9]
the authors discussed the accelerated life testing for battery
discontinuing systems (BDS).The acceleration factor was
determined to adjust the time constant of BDS for better
performance. Also the accelerated life tester must be
enhanced to perform life tests for all the power sources
including solar cells, etc. New methods are required to
accelerate this test while achieving valid results like in [10]
the authors discussed the importance of new bench life-test for
batteries (especially for maintenance free batteries) and
provided recommendations for more precise results.
VI.
ACKNOWLEDGMENTS
The authors are grateful to DDG QM Dte. Gen. Dr. Rasheed
Ahmed Baloch for providing the opportunity for research in
the field of accelerated life testing. The authors oblige the help
and support offered by the officials of Power Electronics Lab
and Electronic Calibration Lab, QM Dte. Gen.
REFERENCES
[1]
[2]
[3]
Zhen-Po, W., Peng, L., & Li-Fang, W. (2013). Analysis on the capacity
degradation mechanism of a series lithium-ion power battery pack based
on inconsistency of capacity. Chinese Physics B, 22(8), 088801.
Gu, W., Sun, Z., Wei, X., & Dai, H. (2014). A Capacity Fading Model
of Lithium-Ion Battery Cycle Life Based on the Kinetics of Side
Reactions for Electric Vehicle Applications. Electrochimica Acta, 133,
107-116.
Eom, S. W., Kim, M. K., Kim, I. J., Moon, S. I., Sun, Y. K., & Kim, H.
S. (2007). Life prediction and reliability assessment of lithium secondary
batteries. Journal of Power Sources, 174(2), 954-958..
38
Journal of Space Technology, Volume V, No.1, July 2015
[4]
Onoda, Y., Noda, M., Abe, Y., & Kobayashi, K. (2001, October).
Development of reliable and long life VRLA batteries. In
Telecommunications Energy Conference, 2001. INTELEC 2001.
Twenty-Third International (pp. 12-16). IET.
[5] Steffke, K. W., Inguva, S., Van Cleve, D., & Knockeart, J. (2013).
Accelerated Life Test Methodology for Li-Ion Batteries in Automotive
Applications (No. 2013-01-1548). SAE Technical Paper.
[6] Tsujikawa, T., Matsushima, T., Yabuta, K., & Matsushita, T. (2009).
Estimation of the lifetimes of valve-regulated lead-acid batteries. Journal
of Power Sources, 187(2), 613-619.
[7] Li, T., & Wang, X. (2009, July). Research of accelerated life test of
cycle-life products. In Reliability, Maintainability and Safety, 2009.
ICRMS 2009. 8th International Conference on (pp. 1190-1196). IEEE.
[8] Takei, K., Kumai, K., Kobayashi, Y., Miyashiro, H., Terada, N.,
Iwahori, T., & Tanaka, T. (2001). Cycle life estimation of lithium
secondary battery by extrapolation method and accelerated aging test.
Journal of Power Sources, 97, 697-701.
[9] Kim, S.-M., Lee, K.-M., Cho, J.-H.Accelerated life testings(ALTs) and
acceleration modeling of battery disconnecting systems(BDS)(2013)
Proceedings - 19th ISSAT International Conference on Reliability and
Quality in Design, RQD 2013, pp. 403-407.
[10] Richter, G. (1993). Improvement of bench life-tests for automotive
batteries. Journal of power sources, 42(1), 231-236.
39